From the Editor: Onboard AI


The three co-founders of OnDeck Fisheries AI. From left, Matthew Leighton, Alexander Dungate and Sepand Dyanatkar: OnDeck Fisheries AI.

Could artificial intelligence soon play a substantial role when it comes to counting catch and bycatch onboard commercial fishing vessels?

Well, it might not be right around the corner, but it’s something that appears to be on the horizon.

A startup co-founded by three people — two former University of British Columbia students and one from Dalhousie University — is working on a monitoring system that uses video and artificial intelligence to better calculate catch and bycatch for commercial fishing boats.

The startup, called OnDeck Fisheries AI, is the brainchild of former UBC students Alexander Dungate and Sepand Dyanatkar, along with ex-Dalhousie student Matthew Leighton. Alexander holds a bachelor’s degree in Computer Science + Biology from the University of British Columbia, while Dyanatkar’s bachelor’s degree is in Computer Science and Physics. Leighton’s degree is in Mathematics and Physics. He’s currently pursuing a Ph.D. in Biophysics at Simon Fraser University.

The problem of under reporting bycatch in commercial fisheries is something Daniel Pauly, a renowned fisheries biologist at the University of British Columbia, has studied. Pauly, who sits on the board of advisors for OnDeck Fisheries AI, is among the most published fisheries scientists of all time, according to OnDeck.

He and his team at the Sea Around Us institute at the University of BC have estimated that global commercial catch of fish could be 30% higher than reported, with unreported bycatch being a significant part of the underreporting.

In response to the problem, OnDeck Fisheries AI has developed a video monitoring and scanning system that uses machine learning and “computer vision” to analyze all fish in a net to give quick, accurate estimates of the types and number of fish that have been caught.

If the technology becomes fully formed, it could make for more accurate reporting of both catch and bycatch. One thing that will help in the quest to eventually bring the technology to market is that the Vancouver-BC-based company was recently awarded $10,000 to help it advance its AI concept.

OnDeck was one of four Canadian startups developing oceans technology that each recently received $10,000 prizes in a competition run by the Ocean Startup Project, a venture whose goal is to kickstart and nurture new, high-quality ocean startups in Canada.

OnDeck’s owners have said the funding is going toward continuing to build automated video review software for scalable and affordable electronic monitoring of commercial fisheries.

Their solution could assist observation companies that currently review millions of hours of video footage from fishing vessel decks manually, thereby helping them cut costs and scale.

OnDeck Fisheries AI also recently won a $25,000 (USD) prize through Dempsey Startup Competition hosted by the University of Washington’s Foster School Buerk Center for Entrepreneurship.

So, while the technology that OnDeck Fisheries AI is working on may not yet be ready for mass use, thanks to the prize money the company has recently won, it should just be a matter of time before its technology comes to market.

More about the company can be found at https://www.ondeck.fish.

Managing Editor Mark Nero can be reached at mark@maritimepublishing.com.

Source: https://fishermensnews.com/from-the-editor-onboard-ai/